How Loud Should My Mix Be: Part 1- WTF is Loudness

I remember the world’s greatest band and the time I was in it – Marketing through twins - My first car and how I confused it for a mastering engineer – WTF do you mean "loudness"? – Busta Rhymes Fractal – Ohhh. LOUDness.

The World's Greatest Band

I can still remember being in the world’s greatest band. It was high school. The year was 2000 and 2 - as we used to never say. My band mates were twins (which is GREAT for marketing) and they had some sort of digital recording machine in their basement. I honestly can’t remember what it was. None of us had any idea of what we were doing with it but, audio gods be damned, we were making records.

At one point someone suggested mastering.

“Hey, I think we should get these songs mastered”.

We didn’t even know what mixing was. Uneasily, I ventured:

“What does mastering do?”

“I think it makes things louder”

“Oh. Yeah. Definitely”

I was confused. Here I was driving around in my 1992 Plymouth Caravan and being perfectly happy making things "louder" by using this thing called a “volume knob”. Instead, I could have been sending songs to a mastering engineering to change their volume. It didn’t make sense.

As it turns out, mastering does typically make songs louder (this is just one of the things that mastering does). This a somewhat more complicated process than what I’m going to explain here but here is an easy way to understand it.

WTF is Loudness

  • In digital audio there is a level “ceiling” above which we do not want to cross. You can see this ceiling on any track meter in your DAW. The top of the meter is marked 0dbFS – which stands for “0 decibels full scale”. It’s not super important what this means but just think of it as a scale on the side of a measuring cup. At 0dbFS the measuring cup is completely full. If we try to add more to the measuring cup, it will overflow. In digital audio this means if we try to go above 0dbFS, we will get nasty clipping. In other words, if you had a recorded sound and you kept bringing the fader up and up, the peaks of the sound would eventually cross 0dbFS and you would get some super nasty digital nails on a digital chalkboard stuff. Go ahead, try it. I’ll wait.

Ok I'll post one too: You can really here it around 5 seconds in.

  • Two audio wave forms that have the same peak level, can still have different perceived volume. In other words, even though the peaks reach the same point on our meter, we will hear one as louder than the other one. This is because of the way the brain perceives volume. Our brains actually perceive the average level of the waveform rather than the peak level of the waveform as “loudness”.

Check it out: Waveform A and Waveform B both peak at -1dbFS. Visually, you can kind of see that Waveform A is a little "fatter" than B. This means that its average level is higher. Which then means that A will sound louder than B even though both of them have the same peak level.

Waveform A (top) and Waveform B (bottom) have the same peak level but Waveform A looks a little "fatter" because its average level is higher

You can easily hear this affect as well. Here are the two wave forms pictured above.

  • Now consider: this same concept applies to your mix – which is effectively just two waveforms (one waveform for the left channel and one waveform to the right channel).

  • This is important because if you loaded up a playlist and songs from different artists had different “average” levels, they would all have different loudness levels – which would be super annoying. You wouldn’t want someone to be listening to a Busta Rhymes banger and then have your puny, soft mix come up right after it, would you? Want to add a caption to this image? Click the Settings icon.

This is a Busta Rhymes Fractal

Ohhh LOUDness! Why didn't you say so.

To put this together, if we want to make your mix louder, we can keep the peaks where they are and just bring up the average level.

How do we do that?

By compressing the dynamic range (translation: by reducing the difference between the peaks and the valleys) of the sound through compression or limiting.

Both compression and limiting reduce the dynamic range of a waveform. They do this in one of two ways:

  1. Downward Compression: Bringing the peaks of the waveform down closer to the valleys – which then allows you to turn the whole waveform up as you bring the peaks back up to their original level

  2. Upward Compression: Bringing the valleys up closer to the peaks

Both processes end up increasing the “average volume” of the waveform – which can then make your tracks more “competitively” loud with the Busta Rhymes track – though your flow probably still wasn’t as clean. Sorry bro. Flip mode is the greatest.

That’s great, right? Turn it up!

Well, yes but there are trade-offs. It turns out that dynamic range is actually part of what makes our mix punchy and exciting. So as we reduce dynamic range, we make things louder but we are also at risk of making things less punchy and less natural sounding.

Think about reducing the contrast on your computer monitor. At some point, things just stop looking “clear”.

In Part 2, we'll talk about the infamous "loudness wars", how and why they started decades ago and how and why they have arguably just ended. Then, we'll see why this is so important for those of us mixing and mastering our own tracks and what tools we can use to help us.